Various approaches for energy storage have been tried. Some batteries that are designed for large scale energy storage have smaller cells arranged in series and parallel. For example, some cells are arranged in parallel, and then that unit is arranged in parallel with another similar unit, and so on. This can require the system to have a disconnect and fuse, and to apply some management strategy that occurs at the high level. These systems can be configured so that they are paralleled at an electrical interface, which can make them complicated to parallel.
One problem with such approaches can be that when batteries are paralleled, one must match their voltage characteristics precisely because they are in parallel electrically. This can significantly limit scalability of the system. For example, one may need to use very similar chemistry, or similar cells, or come up with particularized balancing strategies, to manage the different cells within their ranges of operating characteristics. Also, with regard to the individual cells, the system is in a sense limited by its weakest link. That is, if one cell malfunctions this typically renders the whole array of batteries out of service.